Continuous patient monitoring with AI: real-time analysis of video in hospital care settings
Paolo Gabriel, Peter Rehani, Tyler P. Troy, T.R. Wyatt, Michael A. Choma, Narinder Singh
Abstract
Introduction This study introduces an AI-driven platform for continuous and passive patient monitoring in hospital settings, developed by LookDeep Health. Leveraging advanced computer vision, the platform provides real-time insights into patient behavior and interactions through video analysis, securely storing inference results in the cloud for retrospective evaluation. Methods The AI system detects key components in hospital rooms, including individuals' presence and roles, furniture location, motion magnitude, and boundary crossings. Inference results are securely stored in the cloud for retrospective evaluation. The dataset, compiled with 11 hospital partners, includes over 300 high-risk fall patients and spans more than 1,000 days of inference. An anonymized subset is publicly available to foster innovation and reproducibility at lookdeep/ai-norms-2024 . Results Performance evaluation demonstrates strong accuracy in object detection (macro F1-score = 0.92) and patient-role classification (F1-score = 0.98). The system reliably tracks the “patient alone” metric (mean logistic regression accuracy = 0.82 ± 0.15), enabling detection of patient isolation, wandering, and unsupervised movement-key indicators for fall risk and adverse events. Discussion This work establishes benchmarks for AI-driven patient monitoring, highlighting the platform's potential to enhance patient safety through continuous, data-driven insights into patient behavior and interactions.